Site Health
Your site · Applicable SEO checks passed
View issues and fixesGive AI assistants a short, curated guide to your website. Enter your address or your sitemap and Crawllyt writes an llms.txt file from your real page titles and descriptions, ready to review, edit and publish.
llms.txt is a voluntary convention. It does not grant or block access to your site; robots.txt does that.
An llms.txt file is a plain text file written in Markdown that sits at the top level of a website, at an address like example.com/llms.txt. It was suggested in 2024 as a simple way for site owners to point large language models (LLMs) at the pages that matter most.
Think of it as a reading list rather than a full map. A sitemap lists every address for search engines; llms.txt names your site, explains it in one sentence, and links to a hand-picked set of pages with a short note on each.
AI assistants answer questions by reading web pages, and they can only take in so much text at once. A short, well-labelled index makes it easier for them to find your pricing page, your documentation or your main services instead of guessing from navigation menus and footers.
Support for llms.txt varies between AI companies, so treat it as a low-effort extra rather than a guaranteed way to be cited. It takes minutes to publish and does not affect your normal search results.
# Example Bakery
> Small-batch sourdough bakery in Pune with online ordering and local delivery.
## Main pages
- [Order online](https://example.com/order): Choose loaves and a delivery slot
- [Our story](https://example.com/about): Who bakes your bread and how
## Recipes
- [Starter guide](https://example.com/recipes/starter): Keep a sourdough starter healthy
## Optional
- [Privacy policy](https://example.com/privacy): How customer data is handledNo. robots.txt tells crawlers which addresses they may visit. llms.txt is an optional guide that highlights useful pages. It cannot block or allow anything.
No. It is not a Google Search ranking factor. Its purpose is to help AI tools understand your content when they read your site.
Include the pages a new visitor or an AI assistant most needs: key services or products, pricing, guides and contact details. For most small and medium sites that is roughly 10 to 100 links.
Yes. The result appears in an editable box. Change any title or note before you download it, and keep it up to date as your site changes.
Some sites also publish llms-full.txt, a much longer file with the full text of their pages. Crawllyt creates the shorter llms.txt index, which is the file most tools look for first.
Yes. Create a free Crawllyt account to use it, along with the other audit tools.
Real page checks. Affected URLs. Clear next steps.
Your report will show every check, including passed checks and items requiring manual verification. No scores are shown until a website has been fetched successfully.
Your site · Applicable SEO checks passed
View issues and fixesObserved page signals, not measured AI rankings
How it worksAnalysed pages blocked by robots.txt
View robots.txt rulesGoogle-Extended controls AI usage, not Google Search indexing.
Affected occurrences across checked pages
Expand any row for affected items and instructions. “Review” means the result needs context or cannot be confirmed by a public scan.
Pagination also applies to the tracking, schema and checked-link tables below. Internal links count all occurrences; unique destinations are shown separately.
| Analysed URL | Title / length | Meta description / length | Headings & images | Internal links |
|---|
| Page | Google Analytics | Google Tag Manager | Search Console meta |
|---|
Detected IDs do not confirm that tracking fires correctly. No meta tag does not mean Search Console is unverified: DNS, files and other methods may be used.
AI readiness v2 weights ten observed signals: indexability (15), main heading (10), section structure (10), title (10), summary (10), valid structured data (10), visible content (15), language (5), descriptive internal links (10) and image alt attributes (5). Content and link checks allow partial credit. Thresholds are editorial tool guidelines, not search requirements. A score of 100 means all checklist points were met, not guaranteed AI citations.
GEO averages three observed signals: a question/FAQ heading, a recognised entity schema type, and a descriptive internal link. Scores do not measure AI rankings, citation probability, factual accuracy or content authority. llms.txt and tracking codes do not affect these scores.
SEO score is the percentage of applicable on-page/technical check categories passed. Manual-review checks, contextual “few links” checks and redirects are excluded. It is not a SEMrush or Google score. Character targets are your configured editorial guidelines.
Google AI features guidance · Search Console ownership verification
| Requested URL | Status | Final destination / detail |
|---|
Bounded sample audit, not an exhaustive site crawl. External scripts, consent flows and JavaScript-rendered content are not executed. Blocked sites are reported without invented scores.